Senior Network Engineer - AI Infrastructure & High-Performance Computing

Network Engineer - AI Infrastructure & High-Performance Computing

Up to $250K Base + 2x basic in equity + bonus + Much more!

About the Role

My client is building the network backbone for next-generation AI and machine learning infrastructure. This isn't traditional enterprise networking - you'll be designing and deploying the high-speed network fabrics that connect GPU clusters for large-scale training workloads. If you've worked with HPC environments, GPU clusters, or high-performance datacenter fabrics, this role could be a great fit.

What You'll Be Building

• High-performance datacenter fabrics - Spine-leaf topologies optimized for GPU-to-GPU communication using modern protocols
• Physical network infrastructure - Rack-level designs, structured cabling, and acceptance testing in production datacenters
• Lossless Ethernet networks - Configure congestion control mechanisms for zero packet loss in AI training workloads
• Automated provisioning systems - Infrastructure as Code using Ansible, Terraform, and Python to manage network devices
• Performance validation - Throughput testing with iperf3, NCCL benchmarks, and monitoring with Prometheus/Grafana

This Role Is For You If You've:

✓ Experience with High-Performance Workloads

• Worked with GPU cluster networks, HPC environments, or high-throughput computing infrastructure
• Exposure to InfiniBand (EDR/HDR/NDR) or RDMA over Ethernet (RoCEv2) - even in lab/testing environments
• Some understanding of lossless Ethernet concepts - Priority Flow Control (PFC), Explicit Congestion Notification (ECN)
• Interest in high-bandwidth, low-latency networking for compute-intensive applications

✓ Strong Datacenter Fabric Experience

• Deployed spine-leaf architectures in production (doesn't need to be massive scale)
• Configured VXLAN/EVPN with BGP routing for modern datacenters
• Worked with Arista switches OR Mellanox/NVIDIA networking gear (or willingness to learn)
• Built non-blocking fabrics or understand oversubscription for compute workloads

✓ Hands-On Infrastructure Skills

• Experience with physical datacenter implementations - racking equipment, cable management, labeling
• Comfortable with structured cabling - fiber optics, copper, DAC cables
• Can perform acceptance testing - link validation, basic throughput checks
• Willing to travel to datacenters as needed for hands-on work

✓ Network Automation Experience

• Built configurations using Ansible, Terraform, or Python
• Comfortable with command-line interfaces and scripting
• Experience with version control (Git) for network configs
• Familiarity with monitoring tools - Prometheus, Grafana, or similar platforms

Required Technical Skills

Core Networking (Must Have):

• 5+ years working with datacenter networks
• Strong Layer 2/3 networking - BGP, OSPF, VXLAN, EVPN
• Experience with spine-leaf topologies in production
• Understanding of routing protocols and modern datacenter designs

High-Performance Networking (Preferred):

• Experience with one or more of the following:
• InfiniBand networks (any exposure counts)
• RoCEv2 / RDMA networking
• GPU cluster connectivity
• HPC interconnects or high-bandwidth environments
• Basic understanding of congestion control - ECN, PFC, jumbo frames
• Interest in learning about lossless transport for RDMA workloads

Vendor Platforms (Experience with One or More):

• Arista switches - 7000/8000 series preferred, but any Arista experience valuable
• Mellanox/NVIDIA networking - Spectrum switches, ConnectX NICs (even basic exposure)
• Cisco Nexus datacenter switches - 9K/7K series
• Palo Alto firewalls - configuration and policy management

Automation & Tooling (Must Have at Least Two):

• Infrastructure as Code - Ansible, Terraform, or Python scripting
• Version control - Git for managing network configurations
• Linux familiarity - comfortable in command-line environments
• Monitoring tools - Prometheus, Grafana, SolarWinds, or similar
• DCIM platforms - NetBox, Device42, or asset management tools

Nice to Have (Bonus Skills)

• InfiniBand experience - any hands-on work with IB fabrics
• RoCEv2 or RDMA - configuration or testing experience
• GPU cluster networking - NVIDIA NVLink, GPUDirect
• Performance testing tools - iperf3, NCCL tests, network benchmarking
• Bare metal provisioning - PXE boot, Redfish/IPMI
• Cloud networking - AWS, Azure hybrid connectivity
• Multi-tenant environments - namespace isolation, traffic segmentation

Day-to-Day Responsibilities

• Design and deploy spine-leaf fabrics for AI cluster deployments
• Configure Arista or Mellanox switches with VXLAN/EVPN overlays
• Implement network configurations for GPU cluster connectivity
• Perform rack-level implementations - cable routing, labeling, testing
• Validate network performance using standard testing tools
• Automate network provisioning using Ansible/Terraform
• Monitor network health and troubleshoot performance issues
• Work with datacenter teams on infrastructure requirements
• Implement firewall policies for network segmentation
• Document network designs and maintain topology diagrams
• Travel to datacenters as needed (up to 20% travel)

Compensation & Benefits

• Base Salary: Competitive, up to $200K depending on experience
• Full benefits package - health, dental, vision, 401(k)
• Relocation assistance available if needed
• Professional development budget for certifications and training
• Opportunity to work with cutting-edge AI infrastructure

What You'll Get

✓ Modern technology - Work with 400G networking, GPU architectures, automation tools

✓ High-impact work - Networks you build enable AI research and breakthroughs

✓ Greenfield deployments - Design networks without legacy constraints

✓ Technical focus - Engineering-first culture, minimal bureaucracy

✓ Career growth - Become an expert in AI infrastructure networking

✓ Learning opportunities - Hands-on with InfiniBand, RDMA, GPU networking technologies

How to Apply

If you have datacenter networking experience and interest in HPC/GPU/high-performance computing, I'd love to hear from you.